The study, released by Principled Technologies (PT), highlights the performance of the Supermicro H14 Hyper DP server compared to a 4-year-old server.

SAN JOSE, CA, UNITED STATES, December 5, 2025 /EINPresswire.com/ -- Organizations of all sizes are considering in-house artificial intelligence (AI) chatbots that combine a large language model (LLM) with their own data, offering the opportunity to improve operations without risking the exposure of private data. Chatbots that utilize retrieval augmented generation (RAG) deliver responses that are especially accurate and current. For these organizations, it’s critical to choose a server solution that can deliver quick response times for multiple users conversing with the LLM simultaneously.

Third party Principled Technologies compared the LLM performance of a new Supermicro H14 Hyper DP server powered by AMD EPYC 9965 processors to that of a 4-year-old server. They recently released a test report with their findings.

“Imagine that the leaders of a company have decided to run an in-house AI chatbot using its own private data combined with RAG,” explains the test report. “They have a 4-year-old Supermicro H12 Ultra server powered by earlier processors in house, but the IT team suspects it is not up to the task and are wondering about an alternative. We used an end-to-end chatbot benchmark service called PTChatterly to explore the capabilities of this older server and a new Supermicro H14 Hyper DP server powered by AMD EPYC 9965 processors. In our tests, the chatbot utilized the Llama 3.2-3B-Instruct large language model [LLM] augmented by RAG with local data. For a server solution to support a given number of simultaneous users, the chatbot had to deliver a complete response to a majority of users within 10 seconds, though answers begin to appear in less than 1 second, so the response time feels much faster.”

The report continues: “The new Supermicro H14 Hyper DP server supported 18 simultaneous users posing a sequence of related questions. In most settings, only a fraction of employees would be asking questions of the chatbot at once, so this server is likely to comfortably support far more users in practice.”

To learn more, read the full report at https://facts.pt/5kPexfD

About Principled Technologies, Inc.
Principled Technologies, Inc. is the leading provider of technology marketing and learning & development services.

Principled Technologies, Inc. is located in Durham, North Carolina, USA. For more information, please visit www.principledtechnologies.com.

Sharon Horton
Principled Technologies, Inc.
[email protected]
Visit us on social media:
LinkedIn
Facebook
YouTube
X

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.

Information contained on this page is provided by an independent third-party content provider. Frankly and this Site make no warranties or representations in connection therewith. If you are affiliated with this page and would like it removed please contact [email protected]